Published online November 1, 2007
PEDIATRICS Vol. 120 No. 5 November 2007, pp. 1223-1224 (doi:10.1542/peds.2007-2375)

This Article Extract Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services E-mail this article to a friend Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My File Cabinet Download to citation manager Request Permissions Citing Articles Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by Lavin, P. T. Search for Related Content PubMed PubMed Citation Articles by Lavin, P. T. Related Collections Premature & Newborn Social Bookmarking                         What's this?

LETTER TO THE EDITOR

Meta-Analysis Combining 2 Previously Reported Trials on Respiratory Distress Syndrome in Neonates

Philip T. Lavin, PhD
Averion International Corp
Southborough, MA 01772

To the Editor.—

The recent Moya et al article¹ reported on a meta-analysis as opposed to patient-level analysis despite clear access to the databases as evidenced by the publication authorships.

The results were misleading for the following reasons.

1. There were clear differences between the Safety and Effectiveness of Lucinactant Versus Exosurf in a Clinical Trial of Respiratory Distress Syndrome in Premature Infants (SELECT)² and Surfaxin Therapy Against Respiratory Distress Syndrome (STAR)³ populations with respect to birth weight (36 g mean lower in the STAR study), gestational age (1 week mean lower in the STAR study), prenatal steroid use (8% higher in the STAR study), and Apgar score (1 unit higher in the STAR study). This is best addressed by performing a combined analysis (at the neonate level) to control for these factors.
   
2. The different randomization allocations were a weak reason to not pool the data. The use of a meta-analysis is not efficient when all such baseline data are available and there are study and treatment imbalances.
   
3. The authors did not state if 1-year adjusted survival was the primary efficacy end point or how the analysis was to be performed. A logistic regression model is advised for a 1-year survival end point to diffuse the impact of statistical test choice (log rank versus Wilcoxon-Gehan, which are well known to handle death timing in different manners).
   
4. The life tables shown in their Fig 2¹ are quite different for the 2 studies. The Fig 2 labels are misleading relative to the text and the contents; Fig 2A looks like the pooled life table without colfosceril. The authors did not cite or explain the clear survival differences between the studies.
   
5. The survival rate with poractant seemed more favorable than that with the other 2 animal-based surfactants, but the authors still tried to make a comparison of combined animal-based surfactants versus lucinactant without any justification. They incorrectly assumed that all animal-based surfactants can be pooled; they also did not examine lucinactant variability across the 2 studies. Thus, this conclusion was without appropriate methodologic basis.
   
6. Numerically, the 1-year survival rate with poractant (21.9% without imputation) is, in fact, the numeric best of the 4 therapies (22.6% averaging nonimputed lucinactant across studies); a patient-level multivariate analysis is recommended to adjust for the study and treatment-group imbalances.
   
7. The STAR authors' claim that the 1-year "corrected" lucinactant was superior to poractant (P = .04) is, at best, a posthoc comparison, because the retrospective power is only 29% to rule out a 5% disadvantage even using the 651 neonates on lucinactant versus the 128 neonates on poractant.
   
8. The STAR study itself only had 23% power to rule out a 5% disadvantage before the study started (unlikely that the study was planned to detect this disadvantage); however, it did have 80% power to rule out a 9.3% retrospective disadvantage but was marginally significant to rule out a 5% disadvantage.
   

The publication, therefore, was misleading. The claims were exaggerated beyond sound statistical analysis principles.

FOOTNOTES

Financial Disclosure: Dr Lavin acts as a consultant for Chiesi Farmaceutici SpA.

REFERENCES

1. Moya F, Sinha S, Gadzinowski J, et al. One-year follow-up of very preterm infants who received lucinactant for prevention of respiratory distress syndrome: results from 2 multicenter randomized, controlled trials. Pediatrics. 2007;119(6) . Available at: www.pediatrics.org/cgi/content/full/119/6/e1361
2. Moya FR, Gadzinowski J, Bancalari E, et al. A multicenter, randomized, masked, comparison trial of lucinactant, colfosceril palmitate, and beractant for the prevention of respiratory distress syndrome among very preterm infants. Pediatrics. 2005;115 :1018 –1029[Abstract/Free Full Text]
3. Sinha SK, Lacaze-Masmonteil T, Valls I, et al. A multicenter, randomized, controlled trial of lucinactant versus poractant alfa among very premature infants at high risk for respiratory distress syndrome. Pediatrics. 2005;115 :1030 –1038[Abstract/Free Full Text]

---

PEDIATRICS (ISSN 1098-4275). ©2007 by the American Academy of Pediatrics

CiteULike    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

This Article Extract Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services E-mail this article to a friend Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My File Cabinet Download to citation manager Request Permissions Citing Articles Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by Lavin, P. T. Search for Related Content PubMed PubMed Citation Articles by Lavin, P. T. Related Collections Premature & Newborn Social Bookmarking                         What's this?